Problem behaviors such as school misconduct, underachievement, aggression, delinquency, and substance use place youth at an increased risk for school failure, involvement in the criminal justice system, and chronic substance abuse. Researchers in the PRB have developed a program of research focused on this area. One study employs a randomized quasi-experimental design to test the protective effects of exposure to a social skills curriculum, enhanced school environment, and parent education in the special intervention group compared to problem behavior in a comparison group receiving usual education only. Two cohorts totaling 2700 students in 7 middle schools were followed through the 6th, 7th, 8th and 9th grades. Approximately half received the special intervention. In multi-variate analyses association with problem behaving peers, lack of self control, and poor school adjustment have been found to be risk factors, and social competence and positive parenting behaviors have been found to be protective factors. Analyses on smoking have shown that peer affiliation is the product of smoking behavior rather than a cause and that parental involvement, monitoring, and expectations protect against an increase in smoking both directly and indirectly by preventing association with friends who smoke. Analyses of the efficacy of the intervention indicates an protective effect against increased smoking, but not against other problem behaviors. [unreadable] [unreadable] A second randomized trial is underway. This study focuses on the reduction of aggression and violent behavior among high-risk, inner-city middle school students. It tests the effects of a group-mentor delivered intervention that seeks to encourage the adoption of positive possible self concepts, affiliation with pro-social youth, and commitment to school, and involvement in pro-social activities. Parent and teacher components reinforce the youth intervention. Two cohorts of students from two inner city middle schools will be randomly assigned to special intervention or usual education comparison group. Students will be measured at 5 time points, fall and spring of 6th and 7th grades, and once during 8th grade. Pilot testing of the measures and intervention activities has been completed and examination of pilot data is underway.